ANALYSIS OF SURFACE WATER POLLUTION IN THE KINTA RIVER USING MULTIVARIATE TECHNIQUE (Penilaian Pencemaran Air Permukaan di Sungai Kinta Menggunakan Teknik Multivariat)

This study aims to investigate the spatial variation in the characteristics of water quality monitoring sites, identify the most significant parameters and the major possible sources of pollution, and apportion the source category in the Kinta River. 31 parameters collected from eight monitoring sites for eight years (2006-2013) were employed. The eight monitoring stations were spatially grouped into three independent clusters in a dendrogram. A drastic reduction in the number of monitored parameters from 31 to eight and nine significant parameters (P<0.05) was achieved using the forward stepwise and backward stepwise discriminate analysis (DA). Principal component analysis (PCA) accounted for more than 76% in the total variance and attributes the source of pollution to anthropogenic and natural processes. The source apportionment using a combined multiple linear regression and principal component scores indicates that 41% of the total pollution load is from rock weathering and untreated waste water, 26% from waste discharge, 24% from surface runoff and 7% from faecal waste. This study proposes a reduction in the number of monitoring stations and parameters for a cost effective and time management in the monitoring processes and multivariate technique can provide a simple representation of complex and dynamic water quality characteristics.